Steinmann recognized the association of serpentinized peridotites, radiolarian cherts, and spilitic lavas 50 years ago in the Alps. Benson, 30 years ago, emphasized the world-wide nature of serpentines associated with alpine-type mountains. Nearly 20 years ago the present writer pointed out the relationship of serpentines to island arcs, adding weight to the hypothesis that island arcs represent an early stage in alpine-type mountain building. Serpentinized peridotites are probably intruded during the first great deformation of a mountain belt and do not recur in subsequent deformations of the same belt. They are typically found in two belts about 120 miles apart, one on either side of the central axis of most intense deformation, but may also occur irregularly through this zone. This enables one to date orogenies by dating the serpentines and to follow the axis of an ancient orogenic belt in some cases for thousands of miles. On this basis it was pointed out in 1937 that the great orogeny in the Appalachians was in the Ordovician, not at the end of the Paleozoic, and the Caledonian Revolution in Scotland was of the same age, and not Silurian. These views are now widely accepted. The concept that geosynclines (long narrow troughs containing a thick section of clastic sediments, commonly of shallow-water origin) localize mountain building was challenged on the basis that such a feature is not present in island arcs before the first deformation, but normally develops later because of that deformation. This idea has met with strong resistance, but the writer maintains his original stand. For most of this century a magnificent argument has gone on between field geologists who have worked on the peridotites of alpine mountains and laboratory investigators of their chemistry (particularly Bowen). The former stoutly maintains that the evidence indicates that they were intruded in a fluid state, as magmas; and the latter equally forcefully has proved to his own satisfaction that such magmas are not possible. The writer casts his vote with the field geologist and believes that the field evidence takes precedence. Probably there is some factor or constituent missing in the laboratory investigations. In recent years the idea that the serpentines were emplaced as solids has gained much favor. Some are unquestionably so emplaced, but many cannot be (see Hiessleitner on Balkan occurrences), so that this concept fails as a general explanation. At the other extreme we find Bailey and McCallien suggesting that the serpentines of Turkey are submarine lava flows. In the last few years it has been demonstrated that peridotitic material occurs at shallow depth beneath the oceans (10–12 km). Placing it at such shallow depth beneath island arcs at the time of their initial deformation makes things appear much easier for the solid intrusionists. Were it not for the occurrence of concordant sills in the relatively little deformed flanks of the orogenic zone and the lack of an internal fabric suggesting solid flow, the hypothesis would be an attractive one. That peridotites occur on fault scarps on the Mid-Atlantic Ridge does not indicate that this feature is to be interpreted as an alpine type of mountain system. In this case the faults have merely brought the peridotitic substratum high enough to be exposed. In the oldest rocks serpentinized peridotites do not occur in belts but are ubiquitous throughout the whole terrane. The best example of this is MacGregor’s description of the Sebakwian of Southern Rhodesia, but it also seems to hold for the oldest rocks of the Canadian Shield. These rocks seem to represent something similar to the present oceanic crust strongly deformed. Finally it is suggested that many features of suboceanic topography might be the result of uplift caused by serpentinization of peridotite below the Mohorovičić discontinuity brought about by water leaking from the interior of the earth. The reaction is reversible inasmuch as increase in temperature could cause deserpentinization. This uplift and subsequent subsidence could be accounted for by this reaction.